A collection of two datasets – Baroque and Modern images – form the basis for the resulting models. (Image: Mariana Sanche & Leetee Wang, PennDesign, University of Pennsylvania 2018)